Sealed Rose SDK & DFIR Media Forensics Tools

Official open-source developer SDKs and Digital Forensics & Incident Response (DFIR) tools for Sealed Rose — AI Deepfake Detection, Synthetic Media Analysis, and Frame-Level Forensics.

Built for DFIR analysts, SOC teams, threat intelligence researchers, and incident response pipelines investigating:

• Executive & Employee Impersonation (synthetic video/audio in business communications)
• Identity & KYC Fraud (face swap manipulations in onboarding videos)
• Social Engineering & Phishing (generative media campaigns)
• Digital Evidence Verification (frame-by-frame anomaly detection and model attribution)

DFIR teams can instantly triage suspicious media files without an account or API setup using the free web verification engine: 👉 Sealed Rose Deepfake Video & Image Detector

Ideal for incident response scripts, Jupyter forensic notebooks, and automated IR triaging pipelines:

pip install sealedrose

from sealedrose import SealedRose

client = SealedRose()

# Triage a suspicious video file collected during an incident
result = client.verify_video(file_path="evidence/suspicious_exec_call.mp4")

print(f"Verdict: {result.get('verdict')}")           # e.g., "AI Deepfake Detected"
print(f"Confidence Score: {result.get('deepfakeScore')}") # 0.0 to 1.0
print(f"Forensic Indicators: {result.get('indicators')}")

npm install sealedrose

import { SealedRose } from 'sealedrose';

const client = new SealedRose();

// Verify video authenticity via public URL or local stream
const result = await client.verifyVideo({
  url: 'https://incident-evidence.internal/case-2026-092/payload.mp4'
});

console.log(result.verdict);       // "Looks Real" or "AI Deepfake Detected"
console.log(result.deepfakeScore); // Confidence rating

For air-gapped sandboxes or containerized security pipelines:

docker run -it --rm derekgallardo01/sealed-rose-detector:latest

• Frame-Level Artifact Detection: Analyzes compression discrepancies, optical flow discontinuities, facial landmark micro-jitters, and frequency-domain anomalies.
• Generative AI Attribution: Identifies artifacts characteristic of diffusion engines (Midjourney, Stable Diffusion, Flux, Sora, Runway Gen-3, Kling, Google Veo).
• Fast Triaging API: Designed to plug directly into SOAR workflows, SIEM alerting hooks, and ticketing pipelines.

• Web Verification: Sealed Rose Video Forensics
• Image & Face Authenticity: Sealed Rose Image Forensics
• Chrome Extension: Sealed Rose on Chrome Web Store
• Firefox Add-on: Sealed Rose on Firefox Add-ons

This project and its SDKs are licensed under the MIT License — see the LICENSE file for details.